Interleaving Scheme for Medical Image Authentication

Abstract

Recent developments in communication and information technologies have brought new challenges for easier access, storage, processing, and distribution of medical data for health information systems (HIS). In advanced medical applications including telemedicine and telediagnosis there is a need to exchange the information over an insecure network and hence protection of the confidentiality, integrity, and authentication of medical images is a vital issue. Digital watermarking has the potential to contribute important services for data authentication and copyright protection. This paper presents the discrete cosine transform (DCT)-based approach for interleaving electronic patient records (EPR) in the form of text and heart rate signal (ECG) with medical images. To improve the security, the text data and ECG signal are encrypted before interleaving in the frequency domain. Predicative coding techniques such as adaptive delta modulation (ADM) and differential pulse code modulation (DPCM) are applied for ECG signal encryption and compression. The quality of the watermarked image is assessed using peak signal-to-noise ratio and normalised mean square error. The qualitative analysis was carried out by taking opinions from the physicians and radiologist and it was found that the perceptual integrity of the medical image was preserved. The experimental results on different medical image modalities (MRI, MRA, CT) are analysed, which demonstrates the efficiency and transparency of the proposed scheme as compared to the earlier reported methods.